Abstract

The utilization of floodwater resources will produce benefits, but it will also pose risks; therefore, it is necessary to strengthen knowledge regarding risk assessment to minimize negative effects. In the present study, the risk factors for the utilization of floodwater resources in water diversion projects were identified, the index system was constructed, and the fuzziness and randomness of the risk were considered. Assessment was performed with respect to the following three projects: water storage, water conveyance, and water pumping. The specific methods to improve the cloud model are as follows: analytic hierarchy process (AHP) is used to calculate subjective weights, entropy weight method and projection pursuit method are used to calculate objective weights, X-conditional cloud is used to calculate index membership degree, and finally combination weight and membership degree are combined to obtain the risk level of flood resource utilization. The above methodology was applied to the risk assessment of floodwater resources utilization in the Jiangsu Province of the East Route of the South-to-North Water Transfer Project. The risk of floodwater resources utilization in high-flow, normal-flow, and low-flow years was evaluated, and the validity and applicability of the assessment method were verified.

INTRODUCTION

In recent years, with the rapid development of social economy, the contradiction between supply and demand of water resources has intensified. Trans-basin water transfer can alleviate the shortage of water resources to a certain extent; however, the water transfer is generally carried out during a non-flood period. If the water transfer is also carried out during the flood season, floodwater resources can be fully utilized. Taking the East Route of the South-to-North Water Transfer Project as an example, the quantity of water is abundant during the flood season. If floodwater resources can be strengthened and managed, the sustainable utilization of water resources on the East Route of the South-to-North Water Transfer will be improved, and in turn, social and economic development will be promoted. Floodwater is an unconventional water resource with dual attributes, bringing both benefits and risks. To use floodwater resources safely, it is necessary to perform risk analysis for floodwater resources utilization.

Foreign scholars have conducted less research on the use of floodwater resources, instead focusing on flood risk and flood disasters and losses. Brenner et al. (2016) integrated orthophotos into flood risk analysis, which periodically assesses risk loss. Su et al. (2018) proposed a multi-objective optimization design framework for flood risk management based on elastic targets and applied it to the Indonesian Basin. China has conducted more research regarding the risk of floodwater resources utilization. Gao et al. (2005) proposed a framework and quantitative analysis method for risk design analysis and demonstration from the perspective of the safe use of floodwater resources. Wang et al. (2015) designed floods with different frequencies and analyzed the appropriateness of their floodwater resources utilization. Wang et al. (2017) evaluated the potential risks and benefits of water resources in the basin.

At present, research on the risk assessment of floodwater resources utilization focuses mainly on reservoir (Zou et al. 2018), basin (Wu et al. 2016), and regional (Di 2016) floodwater resources. There exists no assessment system for the utilization of floodwater resources in water transfer projects, which is a complicated process with many risk factors restricting and interacting with each other. The risk of floodwater resources utilization includes not only the probability of the occurrence of floodwater resources utilization risk (risk rate), but also the loss or severity of engineering accidents (risk degree). There are differences in water conservation structure, operation characteristics, function in floodwater resources development and utilization, risk value, and the influence on social and economic development of water diversion projects. It is impossible to use simple weighted summation to evaluate the risk of floodwater resources utilization.

In the present study, according to the characteristics of floodwater resources utilization and risk generation in water transfer projects, an index system for risk assessment was constructed. Subjective weight was calculated by the analytic hierarchy process (AHP), objective weight was calculated by entropy weight and projection pursuit methods, and the final weight was obtained by combining the subjective and objective factors. The risk of floodwater resources utilization is a fuzzy concept with no distinct boundaries (Chen 2005); thus, the cloud model was used to calculate the membership degree, and the risk level of floodwater resources utilization was determined according to the maximum membership degree criteria. Taking the Jiangsu section of the East Route of the South-to-North Water Transfer Project as an example, the present study evaluated the risk of floodwater resources utilization during high-flow, normal-flow, and low-flow years, and verified the applicability of the assessment method. The risk assessment theory provided an important technical means for the safe utilization of floodwater resources in water transfer projects.

MATERIALS AND METHODS

Study area

The first phase of the East Route of the South-to-North Water Transfer Project started construction in December 2002 and was completed in 2013. The water supply ranges from northern Jiangsu, northeastern Anhui, southern Shandong, Eastern Jiaodong, and Northern Shandong. At present, there are still areas along the East Route where the water supply cannot be covered. The utilization of floodwater resources can alleviate the problem of water supply shortage to a certain extent, and the development and utilization of floodwater resources is necessary.

Prior to the implementation of the second phase of the East Route of the South-to-North Water Transfer Project, joint dispatch and management measures will be adopted to transport floodwater resources from Jiangsu Province to Beijing, Tianjin, and Hebei, so as to realize the rational utilization and allocation of water resources and maximize the benefits of the first phase of the project. Taking the Jiangsu section of the East Route of the South-to-North Water Transfer Project as an example, the present study evaluated the risk of floodwater resources utilization, providing a theoretical basis for the safe utilization of floodwater resources. According to the lakes, water conveyance routes, and pumping stations along the Jiangsu section of the East Route of the South-to-North Water Transfer Project, a generalization was carried out; the generalized diagram is shown in Figure 1.

Figure 1

Generalization of the Jiangsu section of the East Route of the South-to-North Water Transfer Project.

Figure 1

Generalization of the Jiangsu section of the East Route of the South-to-North Water Transfer Project.

Risk index system and risk calculation

Index system construction

Risk is uncertain and fuzzy. Many complicated risk indices should be stratified, classified, and screened to identify the main risk factors, so as to construct a reasonable risk index system of floodwater resources utilization. According to the previous risk assessment system of floodwater resources utilization and the characteristics of water transfer projects, the risk assessment index system of floodwater resources utilization in water transfer projects was constructed. The assessment system was divided into three layers, and the floodwater resources utilization risk of water transfer projects was taken as the target layer. According to the factors of risk, the criteria layer included risk for three kinds of projects: water storage, water conveyance, and water pumping. The index layer consisted of the specific index of each kind of project risk. For water storage projects, the risk index mainly included the lake and reservoir in the system; for water conveyance projects, the risk index mainly included river channels; and for water pumping projects, the risk index mainly included pump stations and sluices.

Indicator system assessment level and criteria

Criteria for the assessment of risk rate

The criteria for the assessment of the risk rate were obtained based on China's dam safety assessment (Li 2006) in combination with the actual operating conditions of the water transfer project. The risk rate assessment grade was divided into five levels, the specific content of which is shown in Table 1.

Table 1

Criteria for the assessment of the risk rate of floodwater resources utilization in water transfer projects

Risk rate description Assessment value Corresponding risk rate Explanation 
Smaller (0,1] 0.000001 − 0.0001 Risk is rare and extremely unlikely to happen 
Small (1,2] 0.0001 − 0.01 Risk is accidental and does not happen easily 
Medium (2,3] 0.01 − 0.1 Risk is possible and may happen 
Large (3,4] 0.1 − 0.5 Risk is expected and will happen more than once 
Larger (4,5] 0.5 − 0.99 Risk is frequent and will happen many times 
Risk rate description Assessment value Corresponding risk rate Explanation 
Smaller (0,1] 0.000001 − 0.0001 Risk is rare and extremely unlikely to happen 
Small (1,2] 0.0001 − 0.01 Risk is accidental and does not happen easily 
Medium (2,3] 0.01 − 0.1 Risk is possible and may happen 
Large (3,4] 0.1 − 0.5 Risk is expected and will happen more than once 
Larger (4,5] 0.5 − 0.99 Risk is frequent and will happen many times 
Criteria for the assessment of risk degree

The criteria for the assessment of the risk degree of floodwater resources utilization in water transfer projects were obtained with reference to the classification of major quality and safety accidents in water conservation projects (Li 2009), the specific contents of which are listed in Table 2.

Table 2

Criteria for the assessment of the risk degree of floodwater resources utilization in water transfer projects

Influence degree Assessment value Explanation 
Slight (0,1] Risk does not cause significant loss or delay 
Medium (1,2] Risk brings a small loss (within 100,000 RMB) or delay (within 5 days) 
Severe (2,3] Risk of loss can be compensated (within 1 million RMB) or result in a 2-week delay 
Grave (3,4] Risk brings a considerable but compensable loss (less than 10 million RMB) or a 2-month delay 
Catastrophic (4,5] Risk can bring irreparable damage (death of personnel, loss of more than 10 million RMB), or a delay longer than 2 months 
Influence degree Assessment value Explanation 
Slight (0,1] Risk does not cause significant loss or delay 
Medium (1,2] Risk brings a small loss (within 100,000 RMB) or delay (within 5 days) 
Severe (2,3] Risk of loss can be compensated (within 1 million RMB) or result in a 2-week delay 
Grave (3,4] Risk brings a considerable but compensable loss (less than 10 million RMB) or a 2-month delay 
Catastrophic (4,5] Risk can bring irreparable damage (death of personnel, loss of more than 10 million RMB), or a delay longer than 2 months 
Criteria for the assessment of risk

The risk assessment grade of floodwater resources utilization in water transfer projects was determined by risk rate and risk degree, which were divided into five grades. The concrete contents are listed in Table 3.

Table 3

Assessment criteria and ranks of risk for floodwater resources utilization in water transfer projects

Risk category Risk grade Assessment value Explanation 
Lower risk (0,1] Risk can be ignored, and the current status of the project can be maintained 
Low risk II (1,2] Risk is acceptable, and risk monitoring should be strengthened 
Medium risk III (2,3] Risk is accepted, and risk management measures may be required 
High risk IV (3,4] Risk can be tolerated, and measures should be taken to reduce the risk level 
Higher risk (4,5] Risk cannot be accepted, and risk management measures should be taken at any cost to reduce risk to a tolerable limit 
Risk category Risk grade Assessment value Explanation 
Lower risk (0,1] Risk can be ignored, and the current status of the project can be maintained 
Low risk II (1,2] Risk is acceptable, and risk monitoring should be strengthened 
Medium risk III (2,3] Risk is accepted, and risk management measures may be required 
High risk IV (3,4] Risk can be tolerated, and measures should be taken to reduce the risk level 
Higher risk (4,5] Risk cannot be accepted, and risk management measures should be taken at any cost to reduce risk to a tolerable limit 

Method of risk calculation

The product of risk and risk degree is taken as the risk of floodwater resources utilization according to the definition of risk. In the process of utilizing floodwater resources in water transfer projects, water storage, water conveyance, and water pumping projects bear the main engineering risks. The specific methods for the calculation of risk rate and risk degree are discussed below.

Method for the calculation of risk rate

The risk rate was calculated by formulas (1), (2), and (3) shown below; the calculated results were subsequently converted into risk rate assessment values according to the criteria displayed in Table 1.

① Calculation of the risk rate of water storage projects.

Following an increase in the water storage capacity, the flood control ability of a project is reduced, which increases the risk of the water level exceeding the allowable maximum during the flood season. For a certain flood process Q(t), the highest water level that the project may reach is Hm(t), and risk occurs when Hm(t) exceeds the maximum allowable water level Ha. The risk rate R1 of a water storage project can be expressed as follows: 
formula
(1)

② Calculation of the risk rate of water conveyance projects.

The flood storage capacity of water conveyance projects is an important factor that affects the utilization of floodwater resources in water transfer projects. A water conveyance project provides the necessary conditions for floodwater resources utilization in a water transfer project; when the flow Qc exceeds the maximum allowable flow Qcm of the project, risk is produced, and the risk rate of a water conveyance project R2 can be expressed as: 
formula
(2)

③ Calculation of the risk rate of water pumping projects.

Water pumping projects should be operated normally under the design working conditions. The increased floodwater resources cause the running water level of the project to rise; thus, it is necessary to check whether the water level meets the normal operation requirements at the time of water adjustment. When the pumping water level Hp exceeds the engineering design water level Hpm, the risk will be raised. The risk rate R3 of a water pumping project can be expressed as: 
formula
(3)
Method for the calculation of risk degree

Risk degree assessment is a complicated process, and the risk loss is used to measure the severity of risk consequences. For water conservation and water conveyance projects, the inundation loss caused by flood overflow is mainly considered. For water pumping projects, the economic loss caused by the water supply delay is mainly considered. The value of risk assessment should be based on the criteria displayed in Table 2 in combination with specific engineering parameters, protection scope, water transfer, and other factors.

Calculation of the combined weight

The weight of a risk index is calculated by combining subjective and objective weights. The subjective weight is obtained by expert judgment, and the objective weight can be obtained by combining the intrinsic attributes of the index values. In the present study, the subjective weight was calculated by the analytic hierarchy process, the objective weight was calculated using the entropy weight and projection pursuit methods, and the combined subjective and objective weight was solved by a genetic algorithm.

Analytic hierarchy process

The analytic hierarchy process (AHP) is a method of subjective empowerment. The calculation steps are as follows (Saaty 2008):

  • (1)

    Construction of the judgment matrix. Through the expert assessment of each index, the judgment matrix (, where m is the number of assessment objects and n is the number of assessment indices) is obtained, and the eigenvalues and eigenvectors of A are subsequently solved.

  • (2)

    Consistency test. The consistency index (CI) test is used to judge whether the matrix A satisfies the consistency requirement; the larger the CI, the more serious the inconsistency.

  • (3)
    Calculation of the index weights. If the hierarchical structure passes the consistency test, the final weight dk of each factor of the index layer can be obtained, where bkj is the relative weight of each index and cj is the weight of the criteria layer: 
    formula
    (4)

Entropy weight method

The entropy weight method measures the amount of information by the degree of information disorder. The more information the indicator carries, the greater the effect of the indicator on decision making. The smaller the entropy value, the smaller the disorder of the system. The judgment matrix of the entropy weight method is composed of assessment index values, which is an objective weighting method (Meng & Hu 2009). The calculation steps are as follows:

  • (1)

    If there are m assessment objects and n assessment indicators, the index values are normalized by the range method to obtain the initial judgment matrix.

  • (2)
    Calculation of the entropy of each indicator: 
    formula
    (5)
     
    formula
    (6)
  • If
    fij = 0, ln fij is meaningless, and then: 
    formula
    (7)
  • In

    the formula, if fij = 0, then ln fij = 0.

  • (3)
    Calculation of the weight of each indicator: 
    formula
    (8)

Projection pursuit method

The projection pursuit method projects high-dimensional data into low-dimensional subspaces, and uses the projection index function to reflect the possibility of a certain feature structure. The projection direction vector reflects the contribution of each index to the projection value, and the normalized projection vector is used as the weight of the assessment index (Liu et al. 2017). The projection pursuit method is an objective weighting method that can obtain the weight directly from the decision matrix. The calculation steps are as follows:

  • (1)

    The original data of each index is normalized to the initial matrix (, where m is the number of assessment objects and n is the number of assessment indices).

  • (2)
    Construction of a projection index function and projection of the initial index onto the direction vector to obtain the projection value Zi of the index:  
    formula
    (9)
    where a is the projection direction vector, a = (a1, … , an).
  • The expression of the projection function Q(a) is: 
    formula
    (10)
    In which: 
    formula
    (11)
     
    formula
    (12)
  • where S(a) is the standard deviation of Zi; d(a) is the local density of Zi; is the mean of Zi; R is the radius of the local density window, usually ; rij is the distance of two projection values; and is the unit step function, such that when , it takes 1, otherwise it takes 0.

  • (3)
    Construction of a fitness function by optimizing the projection function: 
    formula
    (13)
     
    formula
    (14)

Finally, it is solved by RAGA, and the weight value is taken as the average of the best individuals of 50 generations.

Combined weight method

The combined weight method combines subjective weight with objective weight, which is a scientific and reasonable method (Peng et al. 2006). It is solved by a genetic algorithm, and the expression of the fitness function of the genetic algorithm is: 
formula
(15)
 
formula
(16)
where is the weight of the j-th indicator of the i-th method; is the weight of the randomly generated j-th index; m is the number of assessment methods; and n is the number of assessment indices.

Improved cloud model risk assessment

Cloud model and its digital features

The cloud model, proposed by Li & Liu (2004), is a mathematical model of uncertainty for dealing with qualitative concepts and quantitative descriptions. If U is a domain represented by an exact number and R is its corresponding qualitative concept, then there is a stable random number for any element x in U. This random number is the degree of certainty of the element x to the concept R (Wang & Li 2016).

The numerical characteristics of the cloud can reflect the quantitative characteristics of the qualitative concept, expressed by the expected value (Ex), entropy (En), and super-entropy (He). Ex is the central value of the qualitative domain, En measures the ambiguity and randomness of qualitative concepts, and He measures the uncertainty of entropy (Luo et al. 2009). The normal cloud and digital features are shown in Figure 2.

Figure 2

Normal cloud and digital feature graph.

Figure 2

Normal cloud and digital feature graph.

Cloud generator

The cloud generator is the basic algorithm of the cloud model, including the forward cloud generator, backward cloud generator, and conditional cloud generator (Yang et al. 2015), which can realize the uncertainty transformation of quantitative values and qualitative language.

The forward cloud generator (FCG) generates quantitative cloud droplet xi and membership degree Ui through the digital characteristics Ex, En and He of the cloud. It is a process from qualitative concept to quantitative index, as shown in Figure 3.

Figure 3

Forward cloud generator.

Figure 3

Forward cloud generator.

The backward cloud generator (BCG) converts a certain number of exact numerical values xi into qualitative concepts Ex, En and He, and makes analysis and calculation according to these three numerical characteristics. It is a conversion process from quantitative values to qualitative concepts, as shown in Figure 4.

Figure 4

Backward cloud generator.

Figure 4

Backward cloud generator.

The conditional cloud generator is made up of X cloud generators and Y cloud generators, all of which belong to special forward cloud generators. Ex, En, He, and specific conditions X or Y are known to produce membership degrees. Typically, a single-condition, single-rule generator is generated jointly by the X and Y cloud generators, as shown in Figure 5.

Figure 5

Conditional cloud generator.

Figure 5

Conditional cloud generator.

Normal cloud definition: if x satisfies the condition of , where , and the certainty of x to C satisfies: 
formula
(17)
then the distribution of x in the domain U is called a normal cloud (Shi & Zha 2017).

Calculation of the index membership degree

In the present study, the X-condition cloud generator was used to calculate the membership degree of the risk indicator. The calculation steps were as follows (Ding & Wang 2013):

  • (1)

    Determination of the grade criteria of the assessment index.

  • (2)
    Calculation of the eigenvalues of different grades and solving of the membership degree of each index to different grades. 
    formula
    (18)
  • where Bmin and Bmax represent the lower and upper limits of the index assessment grade, respectively, and k is a constant that can be adjusted according to the fuzzy threshold of the variable. The cloud number characteristics of the ratings are shown in Table 4.

  • Through the X-condition cloud generator, the index value is input, repeated 1,000 times for each sample, and the average value as the membership degree of each level is obtained. The output cloud model membership degree matrix is , and the calculation formula is as follows: 
    formula
    (19)
     
    formula
    (20)
Table 4

Rules for computing the digital features of the cloud model

Risk grades Ex En He 
  0.2 
II   0.2 
III   0.2 
IV   0.2 
  0.2 
Risk grades Ex En He 
  0.2 
II   0.2 
III   0.2 
IV   0.2 
  0.2 
By multiplying the combined weight W by D, the fuzzy subset C of the assessment set can be obtained (Wang et al. 2014). The maximum membership degree criteria are used for the grade assessment: 
formula
(21)

Risk assessment process of floodwater resources utilization

In the present study, the risk assessment of floodwater resources utilization in water transfer projects was divided into six steps; the assessment process is shown in Figure 6.

Figure 6

Flow chart for the risk assessment of floodwater resources utilization.

Figure 6

Flow chart for the risk assessment of floodwater resources utilization.

Step 1: According to the main risk factors, the risk assessment index system of flood resources utilization in water transfer projects is constructed.

Step 2: Calculate the risk of flood resources utilization of each assessment index.

Step 3: The subjective weight is solved by AHP, the objective weight is solved by the entropy weight method and projection pursuit method, and the combination weight is calculated by genetic algorithm.

Step 4: Calculate the membership of risk indicators at each risk level through the cloud model.

Step 5: Find out the risk assessment grade of each index.

Step 6: According to the improved cloud model, the overall assessment results of the index system are obtained.

RESULTS AND DISCUSSION

Construction of the indicator system

Figure 7 shows that the starting point for the utilization of floodwater resources in the Jiangsu section is Hongze Lake, and the risk indicators are identified based on the lakes, rivers, and gates of the water transfer route. The risk of the water storage project is composed of the risks of Hongze and Luoma Lakes, and the lower lake of Nansi Lake; the risk of the water transfer project is composed of the risks of Zhongyun, Xuhong, Bulao, and Hanzhuang Rivers; and the risk of the water pumping project is composed of the risks of Siyang, Sihong, Liu Laojian, Suining, Zaohe, Pizhou, and Liushan pumping stations. The assessment index system is shown in Figure 7.

Figure 7

Risk assessment index system for floodwater resources utilization in the Jiangsu section of the East Route of the South-to-North Water Transfer Project.

Figure 7

Risk assessment index system for floodwater resources utilization in the Jiangsu section of the East Route of the South-to-North Water Transfer Project.

Risk calculation

Typically, the flood season occurs from June to September every year. Considering the actual water demand of Jiangsu Province and the water resource regulation of lakes, floodwater resources utilization and the water diversion period of the East Route are adjusted from mid-July to September. Under different inflow conditions, the risk of floodwater resources utilization is different. The present study evaluates the risk of floodwater resources utilization during high-flow, normal-flow, and low-flow years.

This water transfer project has a spatial span; thus, it is necessary to correctly classify the typical years of risk assessment. The selected typical years should reflect the representative changes in water resources in the basin. The typical year selection method is as follows: according to the water resources bulletin and other data regarding the basin, the initial typical year of floodwater resources utilization (25%, 50%, and 75% frequencies in high-flow, normal-flow, and low-flow years) is determined; the discarded water during the flood season is considered the available amount of floodwater resources, and it is necessary to make appropriate adjustments to the typical year division based on the lake flow. In the present study, the discarded water of Hongze Lake during the flood season was taken as the supply of floodwater resources to the East Route, and the typical year was based on the Hongze Lake Basin. Through the analysis and calculation of the hydrological data from 2003 to 2017 in the Hongze Lake Basin, the representative for the selected high-flow year was 2003, the representative for the normal-flow year was 2010, and the representative for the low-flow year was 2014.

To verify the rationality, frequency analysis of inflow during the flood season (from mid-July to September) of floodwater resources utilization was performed in three typical years of high-flow, normal-flow, and low-flow. The frequency curves of water inflow during the flood season of the high-flow year are shown in Figure 8; the average inflows of Hongze and Luoma Lakes and the lower lake of Nansi Lake were 4,707 m3/s, 974 m3/s, and 1,092 m3/s, respectively. The frequency curves of water inflow during the flood season of the normal-flow year are shown in Figure 9; the average inflows of Hongze and Luoma Lakes and the lower lake of Nansi Lake were 2,922 m3/s, 458 m3/s, and 367 m3/s, respectively. During the flood season of the low-flow year, the average inflow of Hongze Lake was 1,218 m3/s, and there was almost no inflow of Luoma Lake or the lower lake of Nansi Lake. During the same period, the inflow of Hongze Lake was much larger than that of Luoma Lake and the lower lake of Nansi Lake, with the most discarded water. The inflow of Luoma and Nansi Lakes was similar; thus, the typical year can be accurately estimated by taking the Hongze Lake Basin as a standard.

Figure 8

Water frequency curve of the water storage project during the flood season of a high-flow year.

Figure 8

Water frequency curve of the water storage project during the flood season of a high-flow year.

Figure 9

Water frequency curve of the water storage project during the flood season of a normal-flow year.

Figure 9

Water frequency curve of the water storage project during the flood season of a normal-flow year.

According to the actual operation of the East Route of the South-to-North Water Transfer Project and the method for the calculation of risk, based on the full analysis of the hydrological data and engineering parameters of each risk index, the risk rate assessment value could be calculated according to the amount of water that could be dispatched, and the risk degree assessment value could be obtained by estimating risk loss. The product of the two was used as the risk assessment value. The risk assessment values of each index during the different typical years of high-flow, normal-flow, and low-flow are shown in Table 5. The risk value of water storage, water conveyance, and water pumping projects was the largest during the high-flow year; thus, risk avoidance measures should be taken.

Table 5

Calculation of risk

Risk category Risk index High-flow year Normal-flow year Low-flow year 
Risk of water storage projects B1 Hongze Lake C1 7.88 5.78 4.01 
Luoma Lake C2 4.58 3.91 3.15 
Lower lake of Nansi Lake C3 5.08 3.44 2.21 
Risk of water conveyance projects B2 Zhongyun River C4 4.95 4.28 3.45 
Xuhong River C5 3.45 3.18 2.66 
Bulao River C6 3.02 2.59 2.14 
Hanzhuang River C7 3.39 2.96 2.45 
Risk of water pumping projects B3 Siyang pumping station C8 4.58 3.98 3.30 
Sihong pumping station C9 3.72 3.21 2.64 
Liu Laojian pumping station C10 1.58 1.38 1.13 
Suining pumping station C11 1.16 0.99 0.82 
Zaohe pumping station C12 4.25 3.72 3.04 
Pizhou pumping station C13 2.85 2.50 2.01 
Liushan pumping station C14 3.88 3.38 2.74 
Risk category Risk index High-flow year Normal-flow year Low-flow year 
Risk of water storage projects B1 Hongze Lake C1 7.88 5.78 4.01 
Luoma Lake C2 4.58 3.91 3.15 
Lower lake of Nansi Lake C3 5.08 3.44 2.21 
Risk of water conveyance projects B2 Zhongyun River C4 4.95 4.28 3.45 
Xuhong River C5 3.45 3.18 2.66 
Bulao River C6 3.02 2.59 2.14 
Hanzhuang River C7 3.39 2.96 2.45 
Risk of water pumping projects B3 Siyang pumping station C8 4.58 3.98 3.30 
Sihong pumping station C9 3.72 3.21 2.64 
Liu Laojian pumping station C10 1.58 1.38 1.13 
Suining pumping station C11 1.16 0.99 0.82 
Zaohe pumping station C12 4.25 3.72 3.04 
Pizhou pumping station C13 2.85 2.50 2.01 
Liushan pumping station C14 3.88 3.38 2.74 

Calculation of the combined weight

After the experts in relevant fields scored the risk indicators, the judgment matrix was constructed, and the subjective weights were subsequently calculated. Taking the risk assessment value of floodwater resources utilization of each index from 2003 to 2017 as the initial judgment matrix of the entropy weight and projection pursuit methods, the objective weights were calculated based on entropy weight and projection pursuit methods. The genetic algorithm was used to solve the combined weights, the calculation results of which are shown in Table 6. According to the weight values, the change curve of index weight calculated by the different methods can be seen in Figure 10. The combined weight was between the weights calculated by the three methods, and subjective and objective factors were comprehensively considered.

Table 6

Weight calculation for risk assessment

Risk category Risk index AHP Entropy weight method Projection pursuit method Combined weight 
Risk of water storage projects B1 Hongze Lake C1 0.2960 0.0434 0.0381 0.1258 
Luoma Lake C2 0.1136 0.1161 0.0851 0.1049 
Lower lake of Nansi Lake C3 0.1300 0.0458 0.0717 0.0825 
Risk of water conveyance projects B2 Zhongyun River C4 0.1388 0.0937 0.0790 0.1038 
Xuhong River C5 0.0475 0.0554 0.0922 0.0650 
Bulao River C6 0.0823 0.0577 0.0727 0.0709 
Hanzhuang River C7 0.0283 0.0837 0.0718 0.0613 
Risk of water pumping projects B3 Siyang pumping station C8 0.0529 0.0457 0.0618 0.0535 
Sihong pumping station C9 0.0081 0.0555 0.0623 0.0420 
Liu Laojian pumping station C10 0.0119 0.1155 0.1030 0.0768 
Suining pumping station C11 0.0055 0.0445 0.0387 0.0296 
Zaohe pumping station C12 0.0251 0.0921 0.0705 0.0626 
Pizhou pumping station C13 0.0286 0.0572 0.0740 0.0533 
Liushan pumping station C14 0.0314 0.0938 0.0790 0.0681 
Risk category Risk index AHP Entropy weight method Projection pursuit method Combined weight 
Risk of water storage projects B1 Hongze Lake C1 0.2960 0.0434 0.0381 0.1258 
Luoma Lake C2 0.1136 0.1161 0.0851 0.1049 
Lower lake of Nansi Lake C3 0.1300 0.0458 0.0717 0.0825 
Risk of water conveyance projects B2 Zhongyun River C4 0.1388 0.0937 0.0790 0.1038 
Xuhong River C5 0.0475 0.0554 0.0922 0.0650 
Bulao River C6 0.0823 0.0577 0.0727 0.0709 
Hanzhuang River C7 0.0283 0.0837 0.0718 0.0613 
Risk of water pumping projects B3 Siyang pumping station C8 0.0529 0.0457 0.0618 0.0535 
Sihong pumping station C9 0.0081 0.0555 0.0623 0.0420 
Liu Laojian pumping station C10 0.0119 0.1155 0.1030 0.0768 
Suining pumping station C11 0.0055 0.0445 0.0387 0.0296 
Zaohe pumping station C12 0.0251 0.0921 0.0705 0.0626 
Pizhou pumping station C13 0.0286 0.0572 0.0740 0.0533 
Liushan pumping station C14 0.0314 0.0938 0.0790 0.0681 
Figure 10

The index weight variation curve of the different methods.

Figure 10

The index weight variation curve of the different methods.

Risk assessment

The membership degree of each index was generated by the X-condition cloud generator; to reduce the error, the calculations were repeated 1,000 times. The assessment level was obtained according to the principle of maximum membership degree. The calculated results and the risk grade assessment of each index are shown in Tables 79. From these results, it can be seen that during a high-flow year, the higher risk indicators were Hongze and Luoma Lakes, the lower lake of Nansi Lake, Zhongyun River, and Siyang and Zaohe pumping stations; during a normal flow-year, the higher risk indicators were Hongze Lake and Zhongyun River; and during a low-flow year, the risk of all indicators was low.

Table 7

Calculation of the risk degree of floodwater resources utilization during a high-flow year

Risk index Index value Membership degree
 
Risk grade 
II III IV 
Hongze Lake C1 7.88 0.0000 0.0000 0.9936 0.0064 0.0000 III 
Luoma Lake C2 4.58 0.0000 0.0675 0.9325 0.0000 0.0000 III 
Lower lake of Nansi Lake C3 5.08 0.0000 0.0054 0.9946 0.0000 0.0000 III 
Zhongyun River C4 4.95 0.0000 0.0121 0.9879 0.0000 0.0000 III 
Xuhong River C5 3.45 0.0000 0.9677 0.0322 0.0000 0.0000 II 
Bulao River C6 3.02 0.0001 0.9961 0.0038 0.0000 0.0000 II 
Hanzhuang River C7 3.39 0.0000 0.9748 0.0252 0.0000 0.0000 II 
Siyang pumping station C8 4.58 0.0000 0.0722 0.9278 0.0000 0.0000 III 
Sihong pumping station C9 3.72 0.0000 0.8750 0.1250 0.0000 0.0000 II 
Liu Laojian pumping station C10 1.58 0.0670 0.9328 0.0002 0.0000 0.0000 II 
Suining pumping station C11 1.16 0.5444 0.4556 0.0001 0.0000 0.0000 
Zaohe pumping station C12 4.25 0.0000 0.3145 0.6855 0.0000 0.0000 III 
Pizhou pumping station C13 2.85 0.0001 0.9979 0.0020 0.0000 0.0000 II 
Liushan pumping station C14 3.88 0.0000 0.7775 0.2225 0.0000 0.0000 II 
Risk index Index value Membership degree
 
Risk grade 
II III IV 
Hongze Lake C1 7.88 0.0000 0.0000 0.9936 0.0064 0.0000 III 
Luoma Lake C2 4.58 0.0000 0.0675 0.9325 0.0000 0.0000 III 
Lower lake of Nansi Lake C3 5.08 0.0000 0.0054 0.9946 0.0000 0.0000 III 
Zhongyun River C4 4.95 0.0000 0.0121 0.9879 0.0000 0.0000 III 
Xuhong River C5 3.45 0.0000 0.9677 0.0322 0.0000 0.0000 II 
Bulao River C6 3.02 0.0001 0.9961 0.0038 0.0000 0.0000 II 
Hanzhuang River C7 3.39 0.0000 0.9748 0.0252 0.0000 0.0000 II 
Siyang pumping station C8 4.58 0.0000 0.0722 0.9278 0.0000 0.0000 III 
Sihong pumping station C9 3.72 0.0000 0.8750 0.1250 0.0000 0.0000 II 
Liu Laojian pumping station C10 1.58 0.0670 0.9328 0.0002 0.0000 0.0000 II 
Suining pumping station C11 1.16 0.5444 0.4556 0.0001 0.0000 0.0000 
Zaohe pumping station C12 4.25 0.0000 0.3145 0.6855 0.0000 0.0000 III 
Pizhou pumping station C13 2.85 0.0001 0.9979 0.0020 0.0000 0.0000 II 
Liushan pumping station C14 3.88 0.0000 0.7775 0.2225 0.0000 0.0000 II 
Table 8

Calculation of the risk degree of floodwater resources during a normal-flow year

Risk index Index value Membership degree
 
Risk grade 
II III IV 
Hongze Lake C1 5.78 0.0000 0.0001 0.9999 0.0000 0.0000 III 
Luoma Lake C2 3.91 0.0000 0.7352 0.2648 0.0000 0.0000 II 
Lower lake of Nansi Lake C3 3.44 0.0000 0.9692 0.0308 0.0000 0.0000 II 
Zhongyun River C4 4.28 0.0000 0.2920 0.7080 0.0000 0.0000 III 
Xuhong River C5 3.18 0.0000 0.9913 0.0086 0.0000 0.0000 II 
Bulao River C6 2.59 0.0002 0.9992 0.0006 0.0000 0.0000 II 
Hanzhuang River C7 2.96 0.0000 0.9968 0.0032 0.0000 0.0000 II 
Siyang pumping station C8 3.98 0.0000 0.6528 0.3472 0.0000 0.0000 II 
Sihong pumping station C9 3.21 0.0000 0.9893 0.0107 0.0000 0.0000 II 
Liu Laojian pumping station C10 1.38 0.1828 0.8171 0.0001 0.0000 0.0000 II 
Suining pumping station C11 0.99 0.7788 0.2211 0.0000 0.0000 0.0000 
Zaohe pumping station C12 3.72 0.0000 0.8811 0.1189 0.0000 0.0000 II 
Pizhou pumping station C13 2.50 0.0004 0.9991 0.0005 0.0000 0.0000 II 
Liushan pumping station C14 3.38 0.0001 0.9772 0.0227 0.0000 0.0000 II 
Risk index Index value Membership degree
 
Risk grade 
II III IV 
Hongze Lake C1 5.78 0.0000 0.0001 0.9999 0.0000 0.0000 III 
Luoma Lake C2 3.91 0.0000 0.7352 0.2648 0.0000 0.0000 II 
Lower lake of Nansi Lake C3 3.44 0.0000 0.9692 0.0308 0.0000 0.0000 II 
Zhongyun River C4 4.28 0.0000 0.2920 0.7080 0.0000 0.0000 III 
Xuhong River C5 3.18 0.0000 0.9913 0.0086 0.0000 0.0000 II 
Bulao River C6 2.59 0.0002 0.9992 0.0006 0.0000 0.0000 II 
Hanzhuang River C7 2.96 0.0000 0.9968 0.0032 0.0000 0.0000 II 
Siyang pumping station C8 3.98 0.0000 0.6528 0.3472 0.0000 0.0000 II 
Sihong pumping station C9 3.21 0.0000 0.9893 0.0107 0.0000 0.0000 II 
Liu Laojian pumping station C10 1.38 0.1828 0.8171 0.0001 0.0000 0.0000 II 
Suining pumping station C11 0.99 0.7788 0.2211 0.0000 0.0000 0.0000 
Zaohe pumping station C12 3.72 0.0000 0.8811 0.1189 0.0000 0.0000 II 
Pizhou pumping station C13 2.50 0.0004 0.9991 0.0005 0.0000 0.0000 II 
Liushan pumping station C14 3.38 0.0001 0.9772 0.0227 0.0000 0.0000 II 
Table 9

Calculation of the risk degree of floodwater resources utilization during a low-flow year

Risk index Index value Membership degree
 
Risk grade 
II III IV 
Hongze Lake C1 4.01 0.0000 0.6287 0.3713 0.0000 0.0000 II 
Luoma Lake C2 3.15 0.0001 0.9925 0.0074 0.0000 0.0000 II 
Lower lake of Nansi Lake C3 2.21 0.0015 0.9982 0.0003 0.0000 0.0000 II 
Zhongyun River C4 3.45 0.0001 0.9641 0.0358 0.0000 0.0000 II 
Xuhong River C5 2.66 0.0001 0.9991 0.0008 0.0000 0.0000 II 
Bulao River C6 2.14 0.0019 0.9979 0.0002 0.0000 0.0000 II 
Hanzhuang River C7 2.45 0.0004 0.9992 0.0004 0.0000 0.0000 II 
Siyang pumping station C8 3.30 0.0000 0.9825 0.0175 0.0000 0.0000 II 
Sihong pumping station C9 2.64 0.0001 0.9990 0.0008 0.0000 0.0000 II 
Liu Laojian pumping station C10 1.13 0.5632 0.4368 0.0000 0.0000 0.0000 
Suining pumping station C11 0.82 0.9220 0.0780 0.0000 0.0000 0.0000 
Zaohe pumping station C12 3.04 0.0001 0.9954 0.0046 0.0000 0.0000 II 
Pizhou pumping station C13 2.01 0.0040 0.9958 0.0002 0.0000 0.0000 II 
Liushan pumping station C14 2.74 0.0001 0.9987 0.0012 0.0000 0.0000 II 
Risk index Index value Membership degree
 
Risk grade 
II III IV 
Hongze Lake C1 4.01 0.0000 0.6287 0.3713 0.0000 0.0000 II 
Luoma Lake C2 3.15 0.0001 0.9925 0.0074 0.0000 0.0000 II 
Lower lake of Nansi Lake C3 2.21 0.0015 0.9982 0.0003 0.0000 0.0000 II 
Zhongyun River C4 3.45 0.0001 0.9641 0.0358 0.0000 0.0000 II 
Xuhong River C5 2.66 0.0001 0.9991 0.0008 0.0000 0.0000 II 
Bulao River C6 2.14 0.0019 0.9979 0.0002 0.0000 0.0000 II 
Hanzhuang River C7 2.45 0.0004 0.9992 0.0004 0.0000 0.0000 II 
Siyang pumping station C8 3.30 0.0000 0.9825 0.0175 0.0000 0.0000 II 
Sihong pumping station C9 2.64 0.0001 0.9990 0.0008 0.0000 0.0000 II 
Liu Laojian pumping station C10 1.13 0.5632 0.4368 0.0000 0.0000 0.0000 
Suining pumping station C11 0.82 0.9220 0.0780 0.0000 0.0000 0.0000 
Zaohe pumping station C12 3.04 0.0001 0.9954 0.0046 0.0000 0.0000 II 
Pizhou pumping station C13 2.01 0.0040 0.9958 0.0002 0.0000 0.0000 II 
Liushan pumping station C14 2.74 0.0001 0.9987 0.0012 0.0000 0.0000 II 

Finally, the risk assessment grades of floodwater resources utilization during three typical years of high-flow, normal-flow, and low-flow were calculated using Equation (21). The final assessment grade was obtained by the maximum membership criteria, the results of which are shown in Table 10.

Table 10

Risk assessment of floodwater resource utilization on the Eastern Route of the South-to-North Water Transfer Project during different typical years of water flow

Typical year Membership degree
 
Risk grade 
II III IV 
High-flow year 0.0213 0.4536 0.5243 0.0008 0.0000 III 
Normal-flow year 0.0371 0.7045 0.2584 0.0000 0.0000 II 
Low-flow year 0.0711 0.8762 0.0527 0.0000 0.0000 II 
Typical year Membership degree
 
Risk grade 
II III IV 
High-flow year 0.0213 0.4536 0.5243 0.0008 0.0000 III 
Normal-flow year 0.0371 0.7045 0.2584 0.0000 0.0000 II 
Low-flow year 0.0711 0.8762 0.0527 0.0000 0.0000 II 

According to the assessment results displayed in Table 10, the risk grade of the high-flow year was III, corresponding to medium risk, and that of the normal-flow and low-flow years was II, corresponding to low risk. From the risk grade of specific indicators, it was found that there were grade III risk indicators during the normal-flow year, and grade II or I indicators during the low-flow year, which further indicate that floodwater resources utilization risk during the low-flow year was the lowest. The assessment results of the present study are consistent with the experts' assessment; however, the assessment method used in the present study is more objective and reasonable. The subjective–objective combination weight takes into account the objective attributes of human factors and the index value itself. The cloud model considers the fuzziness of risk. The combined weight cloud model theory has good adaptability for solving fuzzy risk assessment problems with randomness.

CONCLUSIONS

The utilization of floodwater resources can alleviate the shortage of water resources in China to a certain extent. The rational utilization of floodwater resources should control the risk within a certain range, and risk avoidance measures can be taken when necessary. From an engineering point of view, the risk index system of floodwater resources utilization in water transfer projects was constructed, and the improved cloud model theory was introduced. The subjective–objective combination weights were calculated by the analytic hierarchy process, entropy weight method, and projection pursuit method. The membership degree was calculated based on X-conditional cloud. The fuzziness and randomness were considered to objectively evaluate the floodwater resources utilization risk of the water transfer project. The calculation and analysis of the Jiangsu section of the East Route of the South-to-North Water Transfer Project show that the project risk of floodwater resources utilization is relatively small, and the utilization of floodwater resources on the East Route will bring huge economic and social benefits to the Beijing–Tianjin–Hebei region. The risk assessment of floodwater resources utilization in water transfer projects has practical significance for the safe use of floodwater resources. The risk assessment method used in the present study provides important theoretical support for the safe use of floodwater resources of flood control projects.

The water transfer engineering system is complex, and there are many factors affecting the risk of floodwater resources utilization. The present study only considered the engineering risk, and the water quality risk and management scheduling risk must be studied in the future.

ACKNOWLEDGEMENTS

This work was financially supported by the National Key Research and Development Program of China (2016YFC0400909); the fundamental Research Funds for the Central Universities (2019B11014); the Fundamental Research Funds for the Central Universities (2019B70414); Postgraduate Research & Practice Innovation Program of Jiangsu Province (SJKY19_0501); the Key Project of Water Conservancy Science and Technique of Hunan Province ([2016]194-21); the University Natural Science Research Project in Anhui Province (KJ2017A898) and 2018 Academic Support Program for Academic Topics in Colleges and Universities in Anhui Province (gxbjZD62).

REFERENCES

REFERENCES
Brenner
C.
,
Meisch
C.
,
Apperl
B.
&
Schulz
K.
2016
Towards periodic and time-referenced flood risk assessment using airborne remote sensing
.
Journal of Hydrology and Hydromechanics
64
(
4
),
438
447
.
doi:10.1515/johh-2016-0034
.
Chen
S. Y.
2005
Variable Fuzzy Set Theory and Method for Water Resources and Flood Control System
.
Dalian University of Technology Press
,
Dalian
,
China
.
Di
Z. L.
2016
The Research of Floodwater Resources Utilization in Chengxihu Flood Storage
.
Master's thesis
,
Hefei University of Technology
,
Hefei, Anhui
,
China
.
Ding
H.
&
Wang
D.
2013
The assessment method of water eutrophication based on cloud model
.
Acta Scientiae Circumstantiae
33
(
1
),
251
257
.
doi:10.13671/j.hjkxxb.2013.01.036 (in Chinese)
.
Gao
B.
,
Wang
Y. T.
&
Hu
S. Y.
2005
Adjustment and application of the limited level of reservoirs during the flood season
.
Advances in Water Science
16
(
3
),
326
333
.
(in Chinese)
.
Li
L.
2006
Dam Risk Assessment and Risk Management
.
China Water and Power Press
,
Beijing
,
China
.
Li
A. H.
2009
Operating Risk Analysis of the Storage Subsystem in the East Route Project of South–North Water Diversion
.
Master's thesis
,
Nanjing Hydraulic Research Institute
,
Nanjing
,
China
.
Li
D. Y.
&
Liu
C. Y.
2004
Study on the universality of the normal cloud model
.
Engineering Science
6
(
8
),
28
34
(in Chinese)
.
Liu
Z. Z.
,
Huang
X. F.
,
Fang
G. H.
,
Huang
X. Q.
&
Chen
Y. Q.
2017
Post benefit assessment of dam reinforcement based on modified cloud model
.
Journal of Water Resources and Water Engineering
28
(
5
),
223
231
.
doi:10.11705/j.issn.1672-643X.2017.05.37 (in Chinese)
.
Luo
Y. Q.
,
Xia
J. B.
&
Chen
T. P.
2009
Network performance comprehensive assessment model based on cloud model and entropy weight
.
Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition)
21
(
6
),
771
775
(in Chinese)
.
Meng
X. M.
&
Hu
H. P.
2009
Application of set pair analysis model based on entropy weight to comprehensive assessment of water quality
.
Journal of Hydraulic Engineering
40
(
3
),
257
262
(in Chinese)
.
Peng
T.
,
Zuo
W. L.
&
Liu
Y. L.
2006
Genetic algorithm for assessment metrics in topical web crawling
. In:
Computational Methods
(G. R. Liu, V. B. C. Tan & X. Han, eds), Springer, Dordrecht, The Netherlands, pp.
1203
1208
.
Saaty
T. L.
2008
Decision making with the analytic hierarchy process
.
International Journal of Services Sciences
1
(
1
),
83
98
.
Shi
X. J.
&
Zha
X. C.
2017
Risk assessment of flood disaster in Ankang City based on cloud model
.
Acta Agriculturae Jiangxi
29
(
1
),
107
111
.
doi:10.19386/j.cnki.jxnyxb.2017.01.21 (in Chinese)
.
Su
H. T.
,
Cheung
S. H.
&
Lo
E. Y.
2018
Multi-objective optimal design for flood risk management with resilience objectives
.
Stochastic Environmental Research and Risk Assessment
32
(
4
),
1147
1162
.
Wang
R. Y.
&
Li
Y.
2016
Risk assessment of urban flood disaster based on entropy-cloud model
.
Water Resources and Power
34
(
9
),
61
63 + 86
(in Chinese)
.
Wang
H.
,
Liu
G. F.
&
Wang
H. M.
2014
Urban extreme rainstorm flood loss based on cloud model
.
Journal of Economics of Water Resources
32
(
2
),
15
18
(in Chinese)
.
Wang
Z. J.
,
Zhu
J. F.
&
Shang
S. W.
2015
Determining the risk-moderate criterion for flood utilization
.
Advances in Water Science
26
(
1
),
27
33
.
doi:10.14042/j.cnki.32.1309.2015.01.004 (in Chinese)
.
Wang
Z. Z.
,
Wang
Y. T.
,
Hu
S. Y.
,
Liu
K. L.
&
Cheng
L.
2017
Theoretical framework of floodwater resources utilization in a basin I: quantitative interpretation
.
Journal of Hydraulic Engineering
48
(
8
),
883
891
.
doi:10.13243/j.cnki.slxb.20161100 (in Chinese)
.
Wu
H. Y.
,
Wang
Y. T.
,
Hu
Q. F.
,
Liu
Y.
&
Cui
T. T.
2016
Flood identification and constraints analysis for floodwater resources utilization in Taihu Lake basin
.
Hydro-Science and Engineering
5
,
1
8
.
doi:10.16198/j.cnki.1009-640X.2016.05.001 (in Chinese)
.
Yang
G.
,
Liu
D. W.
,
Chu
F. J.
,
Peng
H. D.
&
Huang
W. X.
2015
Assessment on risk grade of tunnel collapse based on cloud model
.
Journal of Safety Science and Technology
11
(
6
),
95
101
.
doi:10.11731/j.issn.1673-193x.2015.06.015 (in Chinese)
.
Zou
Q.
,
Hu
X. Y.
&
Zhou
M.
2018
Comprehensive study on floodwater resources utilization of Three Gorges Reservoir II: risk analysis and countermeasures
.
Yangtze River
49
(
4
),
11
16 + 22
.
doi:10.16232/j.cnki.1001-4179.2018.04.003 (in Chinese)
.